Abstract A38: Caveolin-1 as a biomarker of drug response and therapeutic target in association with tyrosine kinase inhibitor treatment in prostate cancer models

We have shown that caveolin-1 (Cav-1) is overexpressed in metastatic prostate cancer (PCa) and that virulent PCa cells secrete biologically active Cav-1. Secreted Cav-1 can be taken up by adjacent PCa and tumor-associated endothelial cells and promote tumor angiogenesis. We investigated the effect of dasatinib and sunitinib on proliferation, tyrosine kinase (TK) phosphorylation, and downstream signaling pathways, including Cav-1 expression and secretion in hormone-refractory PCa cell lines (PC-3 and DU145) in vitro. Dasatinib inhibited proliferation of PC-3 and DU145 at doses from 0.05 to 5.0 μM; sunitinib inhibited their proliferation at doses from 0.2 to 20 μM. Dasatinib and sunitinib treatment yielded a dose-dependent reduction in phosphorylation of PDGFR (Y857), VEGFR2 (Y951), Akt (S473), and Cav-1 (Y14) in PC-3 and DU145 cells relative to that in controls. Further, dasatinib treatment of PC-3 and DU145 cells resulted in reduced phosphorylation of FAK (Y861) and Src (Y416). Sunitinib did not cause a similar effect on FAK in PC-3 and DU145 cells but did cause a reduction in Src phosphorylation of DU145, though not of PC-3 cells. It was interesting that dasatinib and sunitinib treatment of PC-3 and DU145 cells each inhibited the secretion of Cav-1 and caused moderate reduction in cellular Cav-1.

To analyze the role of secreted Cav-1 as a biomarker of drug response and therapeutic target in the context of dasatinib and sunitinib treatment in vivo, we used PC-3 and DU145 subcutaneous xenograft models. PC-3 xenografts were treated with dasatinib (15 mg/kg q.d., p.o.), Cav-1 antibody (Ab; 10 μg, q.o.d., i.p.), or combined dasatinib and Cav-1 Ab, or vehicle alone. DU145 xenografts were treated with sunitinib (10 mg/kg q.d., p.o.), Cav-1 Ab (10 μg, q.o.d., i.p.), combined sunitinib and Cav-1 Ab, or vehicle alone. Treatment with either dasatinib or Cav-1 Ab produced significant tumor regression (P = 0.0072 and 0.0307, respectively) compared with that of vehicle or IgG. Combined dasatinib and Cav-1 Ab produced greater tumor regression than either treatment alone, but the differences did not achieve statistical significance. Similarly, treatment with either sunitinib or Cav-1 Ab induced significant DU145 tumor regression (P= 0.0004 and 0.0016, respectively) compared with that of vehicle or IgG. Combined sunitinib and Cav-1 Ab induced greater but not significantly different tumor regression than either treatment alone. We found it interesting that serum Cav-1 levels were significantly lower in dasatinib-treated mice than they were in vehicle-treated mice (P = 0.0271). Sunitinib treatment also led to lower serum Cav-1 levels than the vehicle did, but these differences were not statistically significant (P = 0.0871). An important note is that serum Cav-1 levels correlated positively with tumor growth (wet weight) in combined vehicle- and dasatinib-treated groups (r = 0.48, P = 0.031) and combined vehicle- and sunitinib-treated groups (r = 0.554, P = 0.0065).

Overall, our results show that dasatinib and sunitinib can each inhibit PCa cell growth both in vitro and in vivo and that growth inhibition is associated with inhibition of TK phosphorylation and downstream signaling and suppression of Cav-1 secretion. We also found that reduced serum Cav-1 was associated with dasatinib and sunitinib treatment and that serum Cav-1 levels correlate positively with tumor growth. Moreover, we showed that systemic administration of Cav-1 Ab led to suppressed tumor growth in PCa xenografts. These results suggest roles for secreted/serum Cav-1 as a biomarker of drug response and as a therapeutic target in association with dasatinib or sunitinib treatment in PCa.

Citation Information: Clin Cancer Res 2010;16(7 Suppl):A38